Direct numerical simulation of installation effects on airfoil noise

To evaluate installation effects on velocity statistics and its influence on farfield noise, three Direct Numerical Simulations (DNS) have been run using the Lattice-Boltzmann Method with the PowerFLOW software on the Controlled-Diffusion (CD) airfoil at a Reynolds number of 150 000 and 8 degrees angle of attack installed in the Universite de Sherbrooke (UdeS) wind tunnel. Differences in setup between these DNS simulations are the addition of voxel refinements and turbulent trips to the simulation setup for better capturing of the jet shear layer downstream of the wind tunnel nozzle lip. Results show that the airfoil boundary layer displacement thickness, momentum thickness and shape factor are slightly increased after jet shear layer refinement due to an increase in mean angle of attack caused by a change in shear layer state. Despite these changes caused by the mixing layer state, maximum Reynolds stress magnitude near the trailing edge of the airfoil was changed by only 6%. This indicates that adjustments to the wall pressure statistics which are relevant to trailing edge noise generation was only marginal. As such, changes to boundary layer statistics had limited impact on far-field noise in the mid-frequency range in this operating state. • Changes to the jet shear layer caused an increase in time-averaged angle of attack. • Stronger adverse pressure gradient were triggered, thickening boundary layer at trailing edge • Peak Reynolds stress magnitude increased marginally by 6% at the trailing edge. • Change in trailing edge noise was slight and was overshadowed by stronger secondary noise sources.